Journal of Vertebrate Paleontology 25(4):950-961, December 2005 
? 2005 by the Society of Vertebrate Paleontology 
THE DIFFICULTIES OF IDENTIFYING FLYING SQUIRRELS (SCIURIDAE: PTEROMYINI) IN 
THE FOSSIL RECORD 
RICHARD W. THORINGTON, JR\ CHAD E. SCHENNUM1-2, LINDSAY A. PAPPAS13, and DIANE PITASSY1 
^Department of Zoology, Smithsonian Institution, Washington, District of Columbia 20560-0108, U.S A., Thoringt@si.edu and 
PitassyD@si.edu; 
^Department of Forensic Science, Virginia Commonwealth University, Richmond, Virginia 23284, U.S.A., Present address: 
Virginia Department of Forensic Science, Richmond, Virginia 23219, U.S.A., chad.schennum@dfs.virginia.gov; 
319 Maxwell Street, Bath, Maine 04530, U.S.A., ldavis@bathpublicschools.com. 
ABSTRACT?Two problems are examined in this paper: (1) the identification of flying squirrels in the fossil record by 
means of their teeth; and (2) the identification of features of the limbs that indicate that the animals were gliders. Dental 
features vary widely among flying squirrels and among other squirrels, and a thorough survey demonstrates that most 
features that have been used to distinguish fossil flying squirrels are also found in some tree squirrels. A review of the 
descriptions of fossil flying squirrels reveals few convincing arguments that these animals actually belong to the Ptero- 
myini and none to support the hypothesis that they were gliding animals. Recent flying squirrels exhibit a number of 
distinguishing morphological features in their carpal and tarsal bones and at the proximal and distal ends of their long 
bones. Some of these morphological structures are obligatory features required for gliding locomotion in squirrels and 
hence are diagnostic of flying squirrels. 
INTRODUCTION 
There is a long-standing disagreement on the relationship of 
flying squirrels to other squirrels. Some paleontologists have ar- 
gued that flying squirrels evolved from paramyids independent 
of other squirrels (Major, 1893; Mein, 1970; de Bruijn and Unay, 
1989). Most others have classified flying squirrels within the Sciu- 
ridae (Simpson, 1945; Hoffmann et al., 1993; McKenna and Bell, 
1997), and recent molecular studies support the view that they 
evolved from one particular group of tree squirrels, long after the 
Sciuridae evolved from paramyids (Mercer and Roth, 2003; Step- 
pan et al., 2004). 
In this paper, we address two problems related to this dis- 
agreement. The first is the problem of identifying fossils that 
belong to the lineage of the Recent Pteromyini, the ancestors of 
modern flying squirrels or their close relatives. The second prob- 
lem is determining whether the fossil animals display morpho- 
logic features that clearly suggest that they were gliders. In the 
paleontological literature, the main focus has been on the first 
problem, and it has been almost completely based on dental 
morphology, for the obvious reason that teeth are all we have for 
most taxa. We propose to examine the criteria that have been 
used to distinguish flying squirrels in the fossil record, and to 
determine if they are supported by the dental morphology of 
Recent squirrels, both flying squirrels and all others. Second, we 
submit a list of postcranial characters that should be retrievable 
and recognizable in the fossil record and that would demonstrate 
that the fossils were derived from squirrels that could actually 
glide. 
Fossils have long been identified as belonging to flying squir- 
rels. Major (1893) described Sciuropterus albanensis (= Al- 
banensia albanensis), comparing it with Pteromys phaeomelas 
(= Aeromys tephromelas), and suggested that several other fos- 
sil squirrels were in fact flying squirrels. He reviewed a number 
of Eocene rodents that showed "analogies" to the dental mor- 
phology of Recent flying squirrels, but he included rodents of 
several different families and even some primates. The feature 
on which he focused was that "all have in common an elegant 
sculpturing of the enamel, which gives often a crenate appear- 
ance to the cusps or crests" (Major, 1893:193). It is clear in his 
writing that Major believed this morphological similarity 
strongly implied phylogenetic affinity, and that he considered 
flying squirrels to have originated from Eocene forms different 
from those that gave rise to the other squirrels. An Eocene origin 
of flying squirrels, independent from the origin of other squirrels, 
has more recently been argued by de Bruijn and Unay (1989), 
based on Oligocene fossils purported to be flying squirrels. It has 
also been suggested that flying squirrels evolved twice from dif- 
ferent groups of tree squirrels (Black, 1972; Hight et al., 1974), 
but this is not supported by morphology (Thorington, 1984; 
Thorington et al., 2002) or molecular studies (Mercer and Roth, 
2003). 
There are now 21 genera of fossil flying squirrels recorded in 
the literature (McKenna and Bell, 1997). Of these, eight genera 
include extant species. This is a surprisingly high diversity of 
fossil flying squirrels. There are fewer genera of fossil tree squir- 
rels (13 genera), although extant tree squirrels (approximately 25 
genera) are more diverse than extant flying squirrels (15 genera) 
(Hoffmann et al., 1993; Thorington et al., 1996). 
McKenna (1962) detailed the dental morphology of all genera 
of extant flying squirrels, documenting the differences among 
them, but it was not his purpose to present features that distin- 
guish flying squirrels from other squirrels. He placed the genera 
of Recent flying squirrels into five groups, as follows. 
1. The Glaucomys group (Glaucomys, Eoglaucomys, Pteromys, 
Olisthomys, and Petaurillus), with the exception of Olistho- 
mys (Petinomys setosus), has a simple low-crowned, upper 
cheek tooth pattern, free of heavy crenulation. 
2. The Iomys group (Iomys) has an enlarged and discrete hypo- 
cone. 
3. The Petinomys group (Aeromys, Petinomys, "Petinomys" vor- 
dermanni, and Hylopetes) has varying degrees of crenulation 
along the occlusal surface. 
4. The Trogopterus group (Trogopterus, Pteromyscus, and Belo- 
mys) has highly cuspidate teeth, with ectolophs having angu- 
lar styles. 
5. The Petaurista group (Petaurista, Aeretes, and Eupetaurus) 
has high-crowned teeth, with cross lophs connecting principle 
crests. These patterns vary in complexity between genera and 
species, but tend to wear flat with age. 
McKenna (1962) demonstrated the diversity of dental patterns 
among flying squirrels (Fig. 1), but the teeth of other squirrels 
950 
THORINGTON ET AL.?FLYING SQUIRRELS IN THE FOSSIL RECORD 951 
FIGURE 1. Upper tooth rows of four species of flying squirrels show- 
ing the diversity of tooth patterns. A, Eoglaucomys fimbriatus (P4-M3 
equals 11.1 mm); B, Petaurista petaurista (P4-M3 equals 15.55 mm); C, 
Trogopterus xanthipes (P4-M3 equals 16.1 mm); D, Aeromys tephrome- 
las (P4-M3 equals 17.3 mm). Scale bars equal 5 mm. 
are equally diverse, making it difficult to distinguish between 
flying and other groups of squirrels. 
Mein (1970) re-analyzed McKenna' s characters and came to 
different conclusions about the supra-generic groupings of flying 
squirrels. He listed the features of each of his groups, but like 
McKenna, it was not his purpose to distinguish between flying 
squirrels and other squirrels. Mein's groups were: 
I. The Glaucomys group consisting of Glaucomys, Eoglauco- 
mys, and Iomys. This combined McKenna' s Glaucomys and 
Iomys groups, while excluding Pteromys. Mein did not ex- 
amine Petaurillus, but he surely would have included it in 
this group because it exhibits the group's diagnostic charac- 
ters. These characters are the combination of smooth 
enamel and absence of lophules, with other common fea- 
tures being absence of the metaloph on M3, no mesostyle, 
presence of an anterior sinuside (anterior valley), and ab- 
sence of an anteroconid. He placed two genera of fossils in 
this group, Cryptopterus and Petauria, and also suggested 
that the group includes the fossil flying squirrels of North 
America. 
II. The Petaurista group, including McKenna' s Trogopterus 
group and Pteromys, consists of the genera Pteromys, 
Trogopterus, Pteromyscus, Belomys, Aeretes, Petaurista, and 
Eupetaurus. It is characterized by the combination of 
smooth enamel and presence of lophules. Other common 
characters are presence of a metaloph on M3 and presence 
of a mesostylid in the form of a crest attached to the meta- 
conid. In this group Mein included the fossil genera Mio- 
petaurista, Forsythia, and Pliopetaurista. 
III. The Petinomys group, which is almost the same as McKen- 
na's, includes Petinomys, Hylopetes, and Aeromys. It is char- 
acterized by pitted enamel and no or few lophules. Other 
common characters are the absence of a metaloph on M3 
and of a hypolophid. In this third group Mein included the 
fossils Blackia and Pliopetes. 
By presenting a different interpretation of groups within the 
crown clade of flying squirrels, Mein raises a further problem in 
recognizing fossil flying squirrels. If there is no single set of 
characters that distinguish the teeth of flying squirrels, one needs 
to determine the several sets of characters that distinguish the 
different groups of flying squirrels, and how each set distin- 
guishes them from other squirrels. If the phylogenetic relation- 
ships within the crown clade are unclear, this becomes a more 
difficult problem. This problem is further documented by Thor- 
ington et al. (2002). In this phylogenetic study based on morpho- 
logical characters, dental characters alone mixed flying squirrels 
and tree squirrels. Flying squirrels formed a monophyletic group 
only when postcranial characters were added to the analysis. 
In a seminal paper, James (1963) noted ten features by which 
the teeth of flying squirrels could be distinguished in the fossil 
record from the teeth of other squirrels. These features have 
been accepted and used extensively since then. Presumably, their 
acceptance is the reason why most subsequent authors have not 
felt obligated to justify their identification of fossils as flying 
squirrels. However, Engesser (1979), Pratt and Morgan (1989), 
and Emry and Korth (1996) raised doubts about the identifica- 
tion of flying squirrels from teeth in the fossil record. We share 
these doubts as illustrated by the similarity in dental morphology 
exhibited by the squirrels in Figures 2 and 3. In this paper we 
review James' (1963) dental characters, identifying which flying 
squirrels exhibit which characters, but also listing tree and 
ground squirrels that exhibit these same characters. We then 
propose a series of postcranial characters that do distinguish 
flying squirrels from other clades of tree and ground squirrels. 
MATERIALS AND METHODS 
We examined skulls and postcranial skeletons of squirrels in 
the United States National Museum (USNM) collection, both 
grossly and under the microscope. The list of specimens exam- 
ined is given in Appendix 1. 
RESULTS 
Dental Characters 
James (1963) noted ten features by which the teeth of flying 
squirrels could be identified in the fossil record and distinguished 
from the teeth of other squirrels. In his words, "The dentitions of 
the Petauristinae can be distinguished from those of the Sciuri- 
nae by the possession of some or all of the following charac- 
ters ..." (James, 1963:88). We list these below, noting in which 
genera of flying squirrels each occurs and whether it also occurs 
in any genus of tree or ground squirrels. This tabulation clearly 
demonstrates that no single one of his dental characters discrimi- 
nate flying squirrels from other squirrels, and that no flying 
squirrel exhibits all of these characters. His descriptions are 
given in quotation marks. 
1. "Occlusal surface of teeth usually heavily crenulated, at least 
in early stages of wear" (James, 1963:88). 
Flying squirrels: Pitting and crenulations are characteristic 
of Hylopetes, Petinomys (Fig. 3A), and Aeromys, as noted by 
952 JOURNAL OF VERTEBRATE PALEONTOLOGY, VOL. 25, NO. 4, 2005 
FIGURE 2. Lower tooth rows of A, flying squirrel, Aeromys tephrome- 
las (p4-m3 equals 12.6 mm); B, Major's (1893) fossil squirrel, Sciurop- 
terus albanensis; (m3 equals 4, ffi2 equals 3.5, ml equals 3, p4 equals 2.5 
mm); and C, tree squirrel, Ratufa affinis (p4-m3 equals 12.8 mm), all 
illustrating similar "elegant sculpturing of the enamel" (Major, 1893: 
193). Scale bars equal 5 mm. 
FIGURE 3. Pitting and crenulation on upper tooth rows of three squir- 
rels. A, Asian flying squirrel, Petinomys vordermanni (P4-M3 equals 5.2 
mm); B, African tree squirrel, Protoxerus stangeri (P4-M3 equals 10.7 
mm); and C, Asian giant tree squirrel, Ratufa bicolor (P4-M3 equals 
15.05 mm). Note lophules on Protoxerus and Ratufa. Scale bars equal 
1 mm. 
McKenna (1962) and subsequently by Mein (1970). We have 
tried to make a distinction between pitted enamel and crenu- 
lated (folded) enamel, but a gradation between the two blurs 
the distinction, particularly in Aeromys. 
Other squirrels: Pitting is prominent on unworn teeth of 
the African squirrels Protoxerus (Fig. 3B) and Heliosciurus. 
Occasional pitting is seen in the African Paraxerus and in the 
Asian genera, on the lower molars of Callosciurus, Lariscus, 
and Prosciurillus. Crenulations are most prominent in the 
Asian giant tree squirrels, Ratufa (Fig. 3C), and some are 
evident on the lower molars of the same three Asian genera, 
Callosciurus, Lariscus, and Prosciurillus. 
2. "Protocone and hypocone distinct, connected by narrow 
loph with deep pit on lingual surface of crown directly below 
loph" (James, 1963:88). 
Flying squirrels: James was probably referring to the pro- 
tocone-hypocone complex of Pteromyscus (Fig. 4A), Belo- 
mys, and Trogopterus. However, there is a distinct hypocone 
in lomys and a small hypocone in Aeromys. In both a narrow 
loph connects the two cones on at least some teeth and there 
is a small to deep depression on the lingual side of the loph. 
Other squirrels: Distinct protocone and hypocone without 
a connecting loph are seen in the Asian ant-eating Rhino- 
sciurus. Small hypocones with a narrowing loph connecting 
them with the protocones are evident in the African Pro- 
toxerus, Heliosciurus, and Paraxerus, in the Asian Callosciu- 
rus (Fig. 4B), Dremomys, and Prosciurillus, and in the Asian 
giant squirrel Ratufa macroura. 
3. "P4 triangular resulting from large, well-developed parastyle 
and reduced anterior cingulum" (James, 1963:88). 
Flying squirrels: A somewhat triangular P4 occurs in Glau- 
comys, Petaurillus, Petinomys, Aeromys, Belomys, and 
Pteromyscus, but in some cases trapezoidal would be a better 
description. The teeth seem best characterized by the reduc- 
tion of the lingual end of the anterior cingulum, which is 
more developed in the other genera of flying squirrels. 
Other squirrels: A similar triangular P4 is common in 
other squirrels, Sciurus and Microsciurus, the African genera 
Heliosciurus and Epixerus, the African ground squirrels At- 
lantoxerus and Xerus, the Indian striped squirrel Funambu- 
lus, the giant squirrels, Ratufa, and the Asian pygmy squirrel, 
Nannosciurus. 
4. "Lophs complicated by presence of protolophules and meta- 
lophules in some genera" (James, 1963:88). 
Flying squirrels: The more complex teeth are found in 
Petaurista, Aeretes, Trogopterus, Belomys, Pteromyscus, Eu- 
petaurus, Pteromys. 
Other squirrels: Lophules are prominent features on the 
teeth of Protoxerus (Fig. 3B), Ratufa (Fig. 3C), Paraxerus, 
and Xerus erythropus. 
5. "Accessory lophs arising from protocone, extending into 
central valley and infrequently into anterior valley" (James, 
1963:88). 
Flying squirrels: Medial mesolophs are seen in Hylopetes 
and Petinomys. 
Other squirrels: A prominent mesoloph occurs in all spe- 
cies of Ratufa. 
6. "Attrition similar to that in Paramyinae, cutting action de- 
veloped between parametaconid and lingual surface of pro- 
tocone, and between entoconid and lingual face of hypo- 
cone" (James, 1963:88). 
Remarks: The "parametaconid" is now regarded as the 
metaconid. We did not examine this character. It seems to 
have been completely neglected in the literature of fossil 
THORINGTON ET AL.?FLYING SQUIRRELS IN THE FOSSIL RECORD 953 
FIGURE 4. A, B, hypocones (arrow) on upper teeth of two squirrels. 
A, Asian flying squirrel Pteromyscus pulverulentus (P4-M3 equals 8.4 
mm); B, Asian tree squirrel Callosciurus prevostii (P4?M2 equals 7.1 
mm). C, D, mesoconids (arrow) on lower teeth of two squirrels. C, 
American flying squirrel Glaucomys sabrinus (p4-m3 equals 8.65 mm); 
D, American tree squirrel Tamiasciurus hudsonicus (p4-m3 equals 8.75 
mm). Scale bars equal 1 mm. 
flying squirrels. Also, the retention of an ancestral character 
is not a good basis for defining or recognizing a derived 
group. 
7. "Mesostyles may be present on P4 to M2" (James, 1963:88). 
Flying squirrels: Seen in most Hylopetes, Petinomys, Aero- 
mys, Petaurista, Trogopterus, Belomys, and Pteromyscus. 
Other squirrels: Present in all specimens examined of Sciu- 
rus, Tamiasciurus, Ratufa, most species of Paraxerus, and 
some specimens of Protoxerus and Paraxerus poensis. 
8. "Mesoconid distinct" (James, 1963:88). 
Flying squirrels: Mesoconids are found in all flying squir- 
rels (Fig. 4C). 
Other squirrels: A distinct mesoconid is visible in Sciurus, 
Tamiasciurus (Fig. 4D), Sciurotamias, Protoxerus, Heliosciu- 
rus, and Ratufa. It is also commonly present in the paramy- 
ids. 
9. "Parametaconid high, entoconid distinct, and metastylid 
present in some specimens" (James, 1963:88). 
Flying squirrels: The metaconid is the highest conid in all 
genera, most distinctively in p4. The entoconid is consis- 
tently the smallest conid, but is distinct in all genera. In 
Iomys, Aeromys, Eoglaucomys, Pteromys, Petinomys geni- 
barbis, Petinomys fuscocapillus, Petinomys lugens, Trogop- 
terus, Belomys, and Pteromyscus, the entoconid is more bul- 
bous in occlusal view. In the remaining genera, the entoconid 
is as wide or only slightly wider than the posterolophid. 
Other squirrels: The combination of these three features is 
found in Sciurus, Tamiasciurus, Ratufa, and Paraxerus. 
10. "Groove present on lingual surface of lower cheek teeth 
opposite connection of protolophid and protoconid" (James, 
1963:88). 
This is the lingual diagonal flexid illustrated by McKenna 
(1962:fig. 5). It is prominent in some species with high- 
crowned teeth, forming between the mesostylid and the en- 
toconid. In other species it is present as a groove, as de- 
scribed by James, or only as a notch, or not at all, and the 
transition from "groove present" to "groove absent" is 
somewhat arbitrary. 
Flying squirrels: Petaurista, Aeretes, Trogopterus, Belomys, 
Pteromyscus, Pteromys, Eupetaurus, Aeromys, Eoglauco- 
mys, Glaucomys, Hylopetes phayrei (but not H. nigripes), 
Petinomys (P. genibarbis, P. lugens, and P. vordermanni, but 
not P. setosus), and Iomys. 
Other squirrels: Ratufa, Protoxerus, Paraxerus vexillarius 
(ml and m2 only), Sciurus (p4 and m2), Tamiasciurus doug- 
lassii (but not T. hudsonicus), Sundasciurus hippurus (m2), 
Marmota, Cynomys, and some Spermophilus (S. beecheyi 
and S. franklinii, but not S. columbianus, S. tridecemlineatus, 
and S. lateralis). 
This documents that no single one of these dental characters 
enables us to distinguish flying squirrels from other squirrels in 
the Recent fauna, and it is doubtful that any would work better 
for identifying flying squirrels in the fossil record. Because a 
combination of features might enable us to do so, we examined 
the combinations used by Mein (1970) to discriminate among his 
three groups of flying squirrels. Mein examined the structure of 
the enamel surface in both fossil and modern flying squirrels, and 
FIGURE 5. A, Proximal and B, distal ends of humeri in three squirrels: 
from left to right, American ground squirrel Spermophilus beecheyi; 
Asian tree squirrel Callosciurus prevostii; and Asian flying squirrel Pe- 
taurista petaurista (scale bars equal 5 mm). See text for distinguishing 
characteristics. 
954 JOURNAL OF VERTEBRATE PALEONTOLOGY, VOL. 25, NO. 4, 2005 
believing this characteristic to be of greatest systematic value, he 
used it to classify the modern flying squirrels into three groups. 
We have listed these groups below and noted which of the non- 
flying squirrels show the characteristics common within each 
group. 
1. Enamel smooth, absence of lophules?Glaucomys, Eoglauco- 
mys, Iomys. 
Other common characters: absence of the metaloph on M3; 
no mesostyle; presence of an anterior sinuside (anterior val- 
ley); absence of anteroconid" (Mein, 1970:51). 
Other squirrels: Smooth enamel and the absence of loph- 
ules occurs in specimens of Atlantoxerus, Epixerus, Funisciu- 
rus, Rubrisciurus, Sciurotamias, Spermophilus, Sundasciurus, 
and Tamiasciurus. These characteristics also occur in some 
species of Callosciurus (C. prevostii), Dremomys (D. everetti), 
Paraxerus (P. poensis, P. boehmi, and P. cepapi), Sciurus (S. 
niger, S. carolinensis, S. stramineus, S. aberti, and S. griseus), 
Tamiops (T. swinhoei), and Xerus (X. rutilus and X. inauris). 
Specimens examined of Exilisciurus, Funambulus, Glyphotes, 
Lariscus, Microsciurus, and Tamiops mcclellandi contained all 
of the characteristics for group one. 
2. Smooth enamel, presence of lophules?Pteromys, Trogop- 
terus, Pteromyscus, Belomys, Aeretes, Petaurista, Eupetaurus. 
Other common characters: presence of a metaloph on M3; 
presence of a mesostylid in the form of a crest attached to the 
metaconid" (Mein, 1970:51). 
Other squirrels: Both the Rhinosciurus and Xerus erythro- 
pus specimens examined showed smooth enamel, lophules, 
and a metaloph on M3. Neither species, however, had the 
mesostylid. 
3. Pitted enamel, no or few lophules?Petinomys, Hylopetes, 
Aeromys. 
Other common characters: no metaloph on M3; no hypo- 
lophid" (Mein, 1970:52). 
Other squirrels: The combination of all four of these fea- 
tures can be found in Callosciurus notatus, Callosciurus fla- 
vimanus, Dremomys pernyi, Heliosciurus, Menetes, Nanno- 
sciurus, Prosciurillus, Ratufa macroura, Sciurus granatensis, 
and Sciurus vulgaris. Other specimens showing some of these 
characteristics include Paraxerus vexillarius, Protoxerus, 
Ratufa affinis, and Ratufa bicolor. 
Clearly, the combinations of characters cited for the Glauco- 
mys group (group one) and the Hylopetes group (group three) do 
not adequately discriminate flying squirrels from other squirrels. 
Mein's Petaurista group (group two) is the most distinctive and 
easily discriminated from other squirrels by their teeth. How- 
ever, the difficulties are compounded if the Hylopetes group is 
combined in the Glaucomys group and if Aeromys is included in 
the Petaurista group, as argued by Thorington et al. (2002) and 
supported by Mercer and Roth (2003). With these latter group- 
ings of genera it becomes difficult, perhaps impossible, to cite 
combinations of dental characters that distinguish the two groups 
from one another and either of them from other squirrels. 
Postcranial Characters 
The complete long bones of flying squirrels are easily distin- 
guished from those of tree squirrels and ground squirrels, be- 
cause they are much longer and more gracile. If complete long 
bones are not available, one can still recognize the long bones of 
Recent flying squirrels by features of their proximal and distal 
ends. In addition, there are trenchant features of carpal and 
tarsal bones that characterize flying squirrels, tree squirrels, or 
ground squirrels. These anatomical features are described below, 
with the presumption that many of them will also distinguish the 
bones of flying squirrels in the fossil record. Unfortunately, post- 
cranial bones of fossil squirrels are seldom found and rarely 
described. If these bones can be identified, as we contend, this 
direct hard evidence will enable us to determine when gliding 
evolved and test the conclusions based on molecular data. 
Humerus?The proximal end of the humerus is distinctive in 
flying squirrels (Fig. 5). In some ground squirrels (Marmota, 
Cynomys, Spermophilus beecheyi) and tree squirrels (Callosciu- 
rus prevostii, C. erythraeus), the deltoid ridge is broad proximally 
and narrows distally. In others it is narrower (Sciurus), but it still 
has obvious lateral and medial edges. In flying squirrels it is less 
prominent, shorter, and narrower still, more like a knife edge. 
The deltoid ridge is more laterally directed in ground squirrels 
than in tree squirrels, and it is even less laterally directed in flying 
squirrels, principally because the ridge is less prominent. We 
were not able to distinguish flying squirrels from other squirrels 
on the basis of the sphericity of the humeral head, the size or 
shape of the greater or lesser tubercles, or the insertion of the 
infraspinatus muscle on the greater tubercle. Tree squirrels and 
ground squirrels can be distinguished by the insertion of the 
infraspinatus, but flying squirrels look like tree squirrels in this 
respect. 
The distal end of the humerus differs among squirrels in sev- 
eral ways (Fig. 5). Tree squirrels and especially ground squirrels 
have a large lateral epicondylar ridge, extending to the lateral 
edge of the epicondyle, for the origin of the brachioradialis and 
the extensor carpi radialis longus and brevis muscles. In most 
flying squirrels (except Eoglaucomys), this ridge is greatly re- 
duced and does not extend to the lateral edge of the epicondyle. 
The medial epicondylar process of all flying squirrels is also rela- 
tively smaller than that of tree or ground squirrels, unlike the 
robust epicondyle of ground squirrels or the more elongate epi- 
condyle of tree squirrels. The distal end of the trochlea is sharply 
angled in tree squirrels, less sharply angled in ground squirrels, 
and in flying squirrels it is even more gradually angled. For this 
character, we find the distinction between tree squirrels and 
ground squirrels to be ambiguous in some cases (e.g., Callosciu- 
rus notatus compared with Spermophilus columbianus), but 
there is no ambiguity between the sharply angled trochlea of tree 
squirrels and the gradually angled trochlea of flying squirrels. 
Between the radial fossa and the medial epicondyle, the shaft 
of the humerus takes a ridge-like form in tree squirrels, but it is 
flat to slightly ridged in ground squirrels. In flying squirrels, this 
ridge is very narrow and prominent. Although we sometimes 
have difficulty distinguishing tree and ground squirrels by this 
character, we found no overlap in appearance between flying 
squirrels and other squirrels. On the extensor surface of the 
humerus, the medial edge of the trochlea is distinctly angled in 
Sciurus, Ratufa, and Callosciurus prevostii. It is much more in 
line with the long axis of the bone in other squirrels, including 
flying squirrels. This character is probably useful only in distin- 
guishing some tree squirrels from flying squirrels. Because there 
is little difference between flying squirrels and Tamiasciurus and 
Callosciurus notatus, it is not safe to use this feature to distin- 
guish flying squirrels from tree squirrels. 
Radius?It is difficult to distinguish different kinds of squirrels 
from the proximal end of the radius. The circumference of the 
proximal articular surface differs slightly in shape, however, and 
ground squirrels have a slight concavity to the circumference on 
the dorsal surface of the radius. This gives the circumference a 
slight kidney-shape, which we have not seen in flying squirrels 
(more circular) and which is rare in tree squirrels (only seen in 
Tamiasciurus). 
At the distal end of the radius, the flexor surface of ground 
squirrels is quite flat. It is round in flying squirrels and many tree 
squirrels (Sciurus, Tamiasciurus, Callosciurus, Ratufa), but it is 
flatter in the African tree squirrels (Protoxerus, Heliosciurus, 
Funisciurus), except for the small species of Paraxerus (P. cepapi 
and P. poensis). 
On the lateral side of the radius, opposite the ulna, a small 
THORINGTON ET AL.?FLYING SQUIRRELS IN THE FOSSIL RECORD 955 
distal process supports the first compartment, for the tendon of 
the abductor pollicis muscle. In flying squirrels, this process is 
more prominent, more pointed, and more separated from the 
styliform process of the radius than it is in other squirrels. 
Ulna?The olecranon process exhibits some subtle differences 
between flying squirrels and other squirrels. In flying squirrels, 
the olecranon is shorter than in tree squirrels and ground squir- 
rels (Fig. 6). Also, the medial edge of the end of the olecranon 
process extends farther medially in ground squirrels, less medi- 
ally in tree squirrels, and least in flying squirrels. The radial 
notch on the ulna has a different orientation relative to the hu- 
meral notch in tree, flying, and ground squirrels. The planes of 
the radial and humeral articular surfaces differ least in the tree 
squirrels, more in the flying squirrels, and most in the ground 
squirrels, which have a sharp angle of separation between the 
two articular surfaces. 
The distal end of the ulna is less robust in flying squirrels, and 
it lacks the distinct groove on the dorsal lateral surface for the 
tendon of the extensor carpi ulnaris muscle. In North American 
tree squirrels (Sciurus) this groove is very prominent. It is less 
prominent in other tree squirrels and in ground squirrels, but 
more prominent than in flying squirrels. On the flexor surface of 
the distal end of the ulna, there is a distinct ridge for the origin 
of the pronator quadratus muscle in tree and ground squirrels. It 
is most prominent in ground squirrels, weaker in tree squirrels, 
and absent in flying squirrels, many of which lack the pronator 
quadratus. 
Carpal Bones?The pisiform bone of flying squirrels is distinct 
from that of tree and ground squirrels. In flying squirrels, the 
pisiform has a strong articulation with the scapholunate bone, as 
well as with the triquetrum and the distal end of the ulna (Thor- 
ington, 1984). This scapholunate process of the pisiform is lack- 
ing in all tree and ground squirrels. In Petaurista, Trogopterus, 
and several other genera of Southeast Asian flying squirrels, 
there is also a distinct triquetral process that extends dorsally 
between the triquetrum and the scapholunate. This is not present 
in the Glaucomys group of flying squirrels. Thus, the shape of the 
pisiform bone is diagnostic of flying squirrels and can be diag- 
nostic of particular groups of flying squirrels (Thorington and 
Darrow, 2000:fig 3; Thorington et al., 2002:fig 9). The scapholu- 
nate bone of flying squirrels is also diagnostic because other 
squirrels lack the distinctive articular facet for the scapholunate 
process of the pisiform. 
Femur?Among Recent squirrels, one can distinguish tree 
squirrels, marmotine ground squirrels, xerine ground squirrels, 
and flying squirrels from the relative sizes and positions of the 
lesser and third trochanters of the femur (Fig. 7). In most tree 
squirrels, the lesser trochanter is large and medial, extending 
medially as far as or farther than the middle of the femoral head. 
It is similarly positioned, although slightly shorter, in many flying 
squirrels (Glaucomys, Eoglaucomys, Iomys), but in spermophil- 
ine ground squirrels it extends posteromedially and usually is 
even shorter. The femoral neck is elongate in the xerine ground 
squirrels, particularly Xerus, causing the lesser trochanter to ap- 
pear relatively much shorter. In one African tree squirrel, Pro- 
FIGURE 6. Proximal ends of ulnae in three squirrels: from top to 
bottom, American prairie dog Cynomys ludovicianus; Asian flying squir- 
rel Aeromys tephromelas; and American tree squirrel Sciurus carolinensis 
(scale bars equal 5 mm). Note differences in shapes and relative lengths 
of the olecranon processes and orientation of radial notches (arrow). 
FIGURE 7. Anterior views of the proximal ends of femora of nine 
squirrels. A, (left to right) flying squirrels Glaucomys volans, Iomys hors- 
fieldii, and Petaurista petaurista; B, tree squirrels Microsciurus alfari, 
Sciurus carolinensis, and Callosciurus prevostii; and C, ground squirrels 
Ammospermophilus leucurus, Spermophilus tridecemlineatus, and Sper- 
mophilus beecheyi (scale bars equal 5 mm). Note differences in size of 
lesser trochanters and third trochanters (arrow). 
956 JOURNAL OF VERTEBRATE PALEONTOLOGY, VOL. 25, NO. 4, 2005 
toxerus stangeri, the lesser trochanter is short. In other African 
tree squirrels it is long (Funisciurus), to very long (Heliosciurus). 
The lesser trochanter is short in the large flying squirrels (Pe- 
taurista, Eupetaurus, Aeromys, Trogopterus) unlike tree squirrels 
and smaller flying squirrels. A prominent third trochanter lies 
opposite or slightly distal to the lesser trochanter in tree squir- 
rels, and is similarly positioned in marmotine ground squirrels. In 
some giant tree squirrels, Ratufa bicolor and Ratufa macroura, 
the third trochanter extends as a prominent ridge, distally and 
slightly posteriorly on the proximal 3/4 of the femur. In xerine 
ground squirrels, the third trochanter is prominent and distal to 
the position of the lesser trochanter. In all flying squirrels, how- 
ever, the third trochanter and the trochanteric ridge are much 
reduced and lie opposite or slightly proximal to the lesser tro- 
chanter. 
The distal end of the femur has few characters that function- 
ally differentiate among squirrels. In large flying squirrels (Pe- 
taurista, Aeromys), however, the patellar groove is broader and 
shallower than in tree, ground, or small flying squirrels. 
Tibia?At the proximal end of the tibia, the popliteal fossa is 
more pronounced in ground squirrels than in tree or flying squir- 
rels (Fig. 8). In ground squirrels, it is limited medially and later- 
ally by low bony ridges, particularly prominent on the medial 
side. In tree squirrels the fossa has distinct edges or low ridges, 
separating it from the lateral surfaces of the bone. In the large 
FIGURE 8. Proximal ends of tibiae of three squirrels, A, anterior view, 
B, posterior view. Left to right, American ground squirrel Spermophilus 
beecheyi, Asian tree squirrel Callosciurus notatus, and Asian flying squir- 
rel Petaurista petaurista (scale bars equal 5 mm). Note differences in 
cnemial crests and popliteal fossae. 
flying squirrels (Petaurista, Eoglaucomys), the edges of the fossa 
are indistinct, with only rounded edges separating the fossa from 
the medial and lateral surfaces of the bone. In the small flying 
squirrels (Glaucomys, Iomys) the fossa is more distinct, similar 
to the condition in tree squirrels. The cnemial crest on the cranial 
surface of the tibia is pronounced in ground squirrels, less pro- 
nounced in tree squirrels, and not very distinct in flying squirrels 
(Fig. 8). 
At the distal end of the tibia, the tibio-fibular articulation is 
long in small flying squirrels (e.g., Glaucomys) and short in large 
flying squirrels. It is relatively short in all tree and ground squir- 
rels. The distal articular surface mirrors the trochlea of the as- 
tragalus. In ground squirrels the medial portion of the articular 
surface is relatively larger than in tree squirrels and flying squir- 
rels. The posterior process at the distal end of the tibia is long in 
tree squirrels, shorter in ground squirrels, and variable among 
flying squirrels, relatively long in Petaurista but short in Glauco- 
mys. In the Glaucomys group, there is a small tubercle on the 
anterior surface at the distal end of the tibia. This is highly di- 
agnostic, because it is the origin of the tibio-carpalis muscle, 
which extends from the wrist to the ankle along the edge of the 
patagium and is found only in flying squirrels (Thorington et al., 
1996:fig. 2; Thorington et al., 2002:fig. 11). The tubercle occurs in 
Glaucomys, Eoglaucomys, Iomys, Hylopetes, Petinomys, and Pe- 
taurillus, but not in the other genera of Recent flying squirrels, in 
which the muscle takes origin from the metatarsals. 
Astragalus?The medial edge of the trochlea for articulation 
with the tibia is much smaller than the lateral edge in tree squir- 
rels and flying squirrels. The two edges are much more similar in 
size in ground squirrels. In flying squirrels, the trochlear surface 
is more 'V-shaped than in other squirrels, to match a sharper 
ridge on the distal articular surface of the tibia. In tree squirrels 
and ground squirrels the middle of the trochlear surface, be- 
tween the medial and lateral edges, is a more rounded groove. 
The neck of the astragalus of tree squirrels (Sciurus, Tamia- 
sciurus, Ratufa, Callosciurus, Sundasciurus) has a small pit just 
distal to the medial side of the trochlear surface. This pit is also 
found in chipmunks (Tamias). It is absent in ground squirrels 
and flying squirrels. It is also absent in African tree squirrels 
(Protoxerus, Heliosciurus, Funisciurus, and Paraxerus), but there 
is a slight groove on the neck at the base of the medial side of the 
trochlear surface in these genera. Because of the 'V-shaped 
groove of the trochlear surface of flying squirrels, the edge of the 
neck adjoining the trochlea is also 'V-shaped. 
On the plantar surface of the astragalus there are two facets 
for articulation with the calcaneus: the proximal astragalo- 
calcaneal articulation and the sustentacular articulation. There is 
commonly a sulcus between the two facets (Fig. 9), within which 
lies the astragalo-calcaneal interosseous ligament and synovial 
tissue of the joints. In some squirrels, the facets are confluent and 
the sulcus is reduced to a pit for the origin of the astragalar end 
of the ligament. The sulcus is completely missing in all genera of 
flying squirrels, but also it is greatly reduced in the North Ameri- 
can tree squirrels (Sciurus, Tamiasciurus), some African tree 
squirrels (Funisciurus, Paraxerus), and some basal members of 
the Marmotini (Tamias, Sciurotamias; the sulcus is very narrow 
in Spermophilus beecheyi and S. variegatus). The sulcus is promi- 
nent in the Asian tree squirrels (Ratufa, Callosciurus, Sunda- 
sciurus), and therefore its presence or absence cleanly distin- 
guishes between Asian tree and flying squirrels. 
Calcaneus?In ground squirrels, the calcaneus is cruciform, 
with the sustentacular process aligned opposite the peroneal pro- 
cess (Fig. 10). In tree and flying squirrels, the sustentacular pro- 
cess is more distal than the peroneal process, so that the two are 
aligned diagonally relative to the main axis of the calcaneus. The 
sustentacular process is triangular in Sciurus and it is elongate in 
Ratufa. It is round, conforming closely to the circular shape of 
the articular process, in ground squirrels (Spermophilus, Cyno- 
THORINGTON ET AL.?FLYING SQUIRRELS IN THE FOSSIL RECORD 957 
FIGURE 9. Plantar surfaces of astragali of five squirrels. A, (left to 
right) tree squirrels Sciurus carolinensis, Callosciurus prevostii, and 
Ratufa bicolor; B, flying squirrels Eoglaucomys fimbriatus and Petaurista 
petaurista (scale bars equal 5 mm). Note presence or absence of sulcus 
(arrows). 
mys, and Xerus) and in some of the Asiatic tree squirrels (Cal- 
losciurus), but not in flying squirrels. There is a distinct sulcus 
between the two articular surfaces for the astragalus in most 
squirrels (including flying squirrels, Ratufa, Callosciurus, Pro- 
toxerus, Heliosciurus, and Funisciurus), but it is practically ab- 
sent in Sciurus and Tamias, and it is small in Spermophilus 
beecheyi and Paraxerus cepapi. Plantar and medial to the cuboid 
articulation there is a prominent process in ground squirrels, 
which is much smaller in flying and tree squirrels. There is always 
a ridge between the peroneal process and the heel in flying squir- 
rels. It is rarely present in tree squirrels (Tamiasciurus) and 
ground squirrels. 
DISCUSSION 
Our understanding of Sciuridae phylogeny and classification 
has changed extensively in the last two years. Two molecular 
studies, each using three genes, six in total, strongly support the 
same phylogeny, with a few minor differences between them 
(Mercer and Roth, 2003; Steppan et al, 2004). Both strongly 
support the monophyly of flying squirrels and place them as the 
sister group of one lineage of tree squirrels, the Sciurini. The 
morphology of modern squirrels is generally concordant with 
this molecular phylogeny. Mercer and Roth (2003) included all 
but one genus of modern flying squirrels and found two main 
groups, a Glaucomys group of six genera and a Petaurista group 
of eight genera. These are the same two groups suggested by 
Thorington and Oarrow (2000) on the basis of ankle and wrist 
morphology and found to be among the most parsimonious 
groupings in a subsequent study of 80 morphological characters 
(Thorington et al., 2002). However, in the Mercer and Roth 
cladogram, two genera in the Petaurista group seem strangely 
placed, particularly with reference to tooth morphology. These 
are Aeretes, which has teeth very similar to Petaurista but is 
placed as a sister group of Trogopterus, and Aeromys placed as 
a sister group of Eupetaurus, in spite of their very different teeth. 
Because Steppan et al. (2004) included only Glaucomys and Pe- 
taurista, their analysis is uninformative about the positions of 
Aretes and Aeromys. 
Mercer and Roth (2003) carefully tested the appropriateness 
of a molecular clock and estimated dates for nodes in their phy- 
FIGURE 10. Dorsal views of calcanei of seven squirrels. A, (left to 
right) ground squirrels Tamias striatus and Spermophilus beecheyi; B, 
flying squirrels Eoglaucomys fimbriatus and Petaurista petaurista; and C, 
tree squirrels Sciurus carolinensis, Callosciurus prevostii and Ratufa bi- 
color (scale bars equal 5 mm). Note cruciform shape in ground squirrels, 
triangular sustentacular process (arrow) in Sciurus, and elongate susten- 
tacular process in Ratufa. 
logeny, providing an hypothesis of flying squirrel phylogeny 
readily compared with the fossil record. If they had used Palaeo- 
sciurus instead of Douglassciurus to root the sciurid tree (Heis- 
sig, 2003), their divergence dates would have been slightly 
younger. They suggest that flying squirrels and the Sciurini (Sciu- 
rus, etc.) diverged in the early Miocene. This is discordant with 
the description of flying squirrels from the Oligocene, in particu- 
lar of Oligopetes described by Heissig (1979) from Germany, and 
the explicit statement by de Bruijn and Unay (1989:144) that 
their observations provide "strong support for the hypothesis of 
Major (1893) and Mein (1970) that the Petauristinae and Sciuri- 
nae do not share a common ancestor." The justification for con- 
sidering Oligopetes to be a flying squirrel is provided by de 
Bruijn and Unay (1989:140): "The complete absence of a proto- 
conule and the near absence of a metaconule in Ml-2, the U- 
pattern formed by the protoloph and the metaloph of the Ml-2, 
and the shape of the occlusal surface of the cheek teeth in Oli- 
gopetes all indicate that we are dealing with true flying squirrels." 
None of these characters truly distinguish flying squirrels from 
other squirrels, most of which lack protoconules and meta- 
958 JOURNAL OF VERTEBRATE PALEONTOLOGY, VOL. 25, NO. 4, 2005 
conules and a number of which exhibit the 'U'-pattern. Further- 
more, Cuenca Bescos and Canudo (1992) illustrated Oligopetes 
with both protoconule and metaconule, and they note the pres- 
ence of these in Oligopetes, Forsythia, Miopetaurista, and Plio- 
petaurista. However, we agree that the combination of characters 
causes the teeth to look remarkably like those of Hylopetes and 
Petinomys. There is approximately a 25 million year gap between 
Oligopetes and Hylopetes, and therefore de Bruijn and Unay 
(1989) did not synonymize the two. 
In contrast, Mercer and Roth (2003) present a case for Hylo- 
petes and Petinomys evolving within the Glaucomys group in the 
mid- to late Miocene, approximately at the time that Hylopetes 
(formerly Pliopetes, synonymized by Bouwens and de Bruijn, 
1986) appeared in the fossil record. Furthermore, in the clado- 
gram, the most basal members of the Glaucomys group are Eo- 
glaucomys and Glaucomys, both of which show a more 'V- 
shaped arrangement of protoloph and metaloph, suggesting that 
ancestral members of this group in the early Miocene would have 
shared this characteristic with the tree squirrels of the same pe- 
riod. 
These differing interpretations of the molecular and paleon- 
tological data may be irresolvable with present evidence. Step- 
pan et al. (2004) included only Glaucomys and Petaurista in their 
analysis and Thorington et al. (2002) were not able to resolve the 
position of Hylopetes and Petinomys with morphological data, so 
neither study helps to support the cladistic placement of these 
two genera by Mercer and Roth (2003). The absence of non- 
dental material of Oligopetes and the long gap in the fossil record 
makes the interpretation of de Bruijn and Unay (1989) less than 
compelling. 
Mercer and Roth (2003) depict the divergence of flying squir- 
rels from the Sciurini in the early Miocene. Early Miocene fossils 
purported to be flying squirrels include the genera Petauris- 
todon, Miopetaurista, Aliveria, and Blackia (McKenna and Bell, 
1992). The genus Sciurion was informally described by Skwara 
(1986) as a "flying squirrel," because it has enamel crenulations 
in the basin of unworn cheek teeth. However, Skwara declined to 
assign it formally to the flying squirrels, because these features 
are found in some paramyids and some species of Prosciurus. 
She allocated it only to the Sciuridae, where it was listed also by 
McKenna and Bell (1997). The identification of Petauristodon as 
a flying squirrel has been questioned by Pratt and Morgan 
(1989), with whom we agree, as elaborated below in our discus- 
sion of the astragalus. Blackia seems to be considered a flying 
squirrel because of its extensively rugose and pitted enamel 
(Mein, 1970). These features have evolved independently in at 
least two lineages of flying squirrels and three of tree squirrels, 
but no tree squirrel exhibits the extreme condition of the enamel 
of Blackia. The simplicity and pattern of Blackia's cusps and 
lophs seem unlike those of any modern flying squirrels, and Mein 
(1970) considered it unlikely to be directly ancestral to any. 
Stronger justification for allocating Blackia to the flying squirrels 
is desirable. 
Aliveria is a reasonable candidate for an early flying squirrel, 
as noted by de Bruijn et al. (1980), and the occurrence of flying 
squirrels at this time is concordant with the molecular data (Mer- 
cer and Roth, 2003). Shuanggouia of the middle Miocene is simi- 
lar to Aliveria and was allocated to flying squirrels because of 
this (Qiu and Liu, 1986). Aliveria and Shuanggouia exhibit the 
convergent protoloph and metaloph of other sciurids with some 
features suggesting they are closely related to Albanensia of the 
middle and late Miocene (McKenna and Bell, 1997) and Mio- 
petaurista. However, the justification for calling these flying 
squirrels seems principally to be the crenulated enamel of the 
talonid basin on some of them. We do not find this convincing 
because it is not characteristic of the basal members of either the 
Glaucomys group (Eoglaucomys, Glaucomys) or the Petaurista 
group (Petaurista). In this respect, interpretation of the fossil 
record is discordant with expectations based on the molecular 
analysis. In clarifying the distinction between Albanensia and 
Miopetaurista, Daxner-Hoch and Mein (1975) did not state why 
they consider these flying squirrels. Albanensia has rugose 
enamel and a strong hypocone and lacks a free mesostyle and 
mesostylid; Miopetaurista has smooth enamel, lacks a hypocone, 
and has a free-standing mesostylid. None of these features are 
diagnostic of flying squirrels; in fact, Petaurista philippinensis 
exhibits a different combination of strong hypocone, smooth 
enamel, and free mesostyle on the Ml-2. Although the large P4 
and the occurrence of lophules on Albanensia are suggestive as 
precursors of the complex high-crowned Petaurista teeth, these 
features might be expected to evolve in common among folivo- 
rous squirrels. 
Forsythia is described (Mein, 1970) as a flying squirrel of the 
middle (?) Miocene, with converging protoloph and metaloph 
and a combination of distinctive 'flying squirrel' features. Mein 
(1970) compared it with Pteromys and Glaucomys, but he did not 
state explicitly why he considered it to be a flying squirrel. He 
may have considered some crenulations on the lower molars to 
be diagnostic. 
Two other genera of flying squirrels have been described from 
the middle Miocene of Asia: Meinia and Parapetaurista. There 
are imprints of the body and a partial skeleton of Meinia, but 
only the teeth were described in English by Qiu (1981). Dimen- 
sions of some of the postcranial elements are also provided in the 
Chinese description, and the illustration suggests that there are 
distal tibia, tarsal bones, and carpal bones. Qiu (1981) considered 
Meinia to be a flying squirrel, closest to Blackia in form and 
pattern of molars and rugosity of the enamel. The figures do not 
persuade us that it is very similar to Blackia, but if it is, it is 
subject to the same questions that we pose for identifying 
Blackia as a flying squirrel. Meinia, however, has postcranial 
elements that may answer these questions. Parapetaurista is com- 
pared by Qiu and Liu (1986) with Miopetaurista, although it 
seems more comparable to Albanensia, because both have ru- 
gose enamel, hypocones, and lack free-standing mesostyles and 
mesostylids. 
Two genera, Pliopetaurista and Pliosciuropterus are reported 
to occur first in the late Miocene (McKenna and Bell, 1992). Qiu 
(1991) cited similarities between Pliopetaurista rugosa and Pe- 
taurista in the general patterns of P4 and M3 and details of DP4. 
Comparing his illustrations with specimens of Petaurista philip- 
pinensis, P. alborufus, and P. xanthipes, we are intrigued by his 
suggestion that Pliopetaurista rugosa is in the lineage leading to 
Petaurista. In contrast, the holotype of Pliopetaurista meini illus- 
trated by Black and Kowalski (1974:plate xv, fig. 1) appears to 
resemble Callosciurus (e.g., C. prevostii) more closely than any 
flying squirrel, especially in the metaloph of Ml-2 with its promi- 
nent metaconule narrowly connected to the metacone and not to 
the protocone. In his description of Pliosciuropterus, Sulimski 
(1964) listed and illustrated the similarities and differences be- 
tween it, Pteromys, and Petaurista. He presents a reasonable, 
though not convincing argument that Pliosciuropterus is a "flying 
squirrel" (his quotation marks). 
Flying squirrels of the Pliocene and Pleistocene are attributed 
to recent genera and to four of the fossil genera discussed above, 
Miopetaurista, Blackia, Pliopetaurista, and Pliosciuropterus. A 
fifth genus, known only as a fossil, is the middle Pleistocene 
Petauria. The teeth illustrated by Dehm (1962) for the type of 
Petauria closely resemble the well-worn teeth of some species of 
Petaurista (c.f. P. philippinensis), and possibly this genus should 
be considered a synonym of Petaurista. If fossils are correctly 
attributed to Recent genera of flying squirrels, there is almost 
certainty that they were gliding animals. Publications detailing 
such attributions of Pliocene and Pleistocene fossils are strongly 
persuasive to us (e.g., Guilday, 1962; Guilday et al., 1964, 1977; 
Chaimanee, 1998). In these publications, reference to postcranial 
THORINGTON ET AL.?FLYING SQUIRRELS IN THE FOSSIL RECORD 959 
elements is rare (Guilday, 1962) even when it is likely that they 
were collected (Guilday et al., 1964, 1977). 
One problem with the attributions of fossils to Recent genera 
is the difficulty of distinguishing between Hylopetes and Petino- 
mys by teeth alone. Bouwens and de Bruijn (1986) concluded 
that it was not possible. This question needs to be revisited, 
because it is unclear what species and specimens they examined. 
They noted that they accepted the names on the specimen labels 
but they did not cite the specimen numbers or their geographic 
origin. The identifications included "Hylopetes saggita" but the 
type specimen of this species is reported to be a Petinomys (Hill, 
i960, 1962; Corbet and Hill, 1992). Also, they listed "Petinomys 
bartelsi" which is now considered to be a Hylopetes (Corbet and 
Hill, 1992; Hoffmann et al., 1993). In addition, there are two 
groups of Hylopetes that can be distinguished by the prominence 
or absence of a mesostyle on P4 and Ml (Thorington et al., 
1996), and Chaimanee (1998) seems to have successfully distin- 
guished between two species of Hylopetes and two species of 
Petinomys from the Pleistocene of Thailand. Therefore it seems 
likely that a careful review of all the species of Hylopetes and 
Petinomys would permit more precise identifications of Pliocene 
and Pleistocene fossils of these squirrels. 
It is probable that Recent flying squirrels are monophyletic 
(Thorington, 1984; Thorington et al., 2002; Mercer and Roth, 
2003), but fossil lineages demonstrated to be gliding squirrels 
need not be closely related to the Recent ones. Gliders have 
evolved among at least seven different lineages of mammals, so 
there is no reason to assume among squirrel fossils that the dem- 
onstration of flying squirrel morphology is also a demonstration 
that the fossil belongs to the Pteromyini. The opposite is also 
possible, that the Pteromyini and the Sciurini diverged before 
gliding evolved in the former. Until there is clear evidence sup- 
porting either of these possibilities, however, we prefer the hy- 
pothesis that gliding morphology evolved once in the Sciuridae 
at the base of the radiation of the Pteromyini. 
In distinguishing among ground squirrels, tree squirrels, and 
flying squirrels in the fossil record, it is desirable to choose fea- 
tures that are functionally important for these different ways of 
life. Diet varies greatly among Recent squirrels, and with a few 
specialized exceptions it varies independent of terrestrial, arbo- 
real, and gliding habits. Many small ground squirrels, tree squir- 
rels, and flying squirrels feed extensively on seeds, insects, and 
other rich sources of nutrition. In contrast, large squirrels feed 
more extensively on leaves and other vegetation. There is no 
reason to expect a tighter correlation among fossil squirrels. This 
does not preclude dental features from being good indicators of 
lineages over geologically short periods of time, but it should 
raise concerns about interpretations of lineages traced by dental 
features over tens of millions of years. This is the crux of our 
concern with identifying flying squirrels in the fossil record solely 
from teeth. Our concern is furthered by our survey of the dental 
features that have been used to distinguish fossil flying squirrels. 
Many of these features appear to be widespread among Recent 
squirrels that are not flying squirrels. We submit that such iden- 
tifications of fossils as flying squirrels should be carefully argued 
with full consideration and exclusion of alternative hypotheses. 
We have not examined the dental features of paramyid ro- 
dents and compared them with the features of flying squirrels, 
because we are not convinced of their relevance. Morphological 
evidence suggests, although not conclusively, that flying squirrels 
are derived from ancestral tree squirrels, not directly from para- 
myids. Flying squirrels have male reproductive tracts that are 
remarkably similar to those of other squirrels. These tracts in- 
clude the bulbar gland and penile duct, which are unique to 
squirrels (Mossman et al., 1932). Unfortunately the primitive 
state of the male reproductive tract in paramyids is unknown. 
Flying squirrels share the sciurimorph jaw musculature (Bryant, 
1945), but sciuromorphy has evolved repeatedly in rodents. Mo- 
lecular data strongly support the origin of flying squirrels from 
tree squirrels (Mercer and Roth, 2003; Steppan et al., 2004), but 
of course, paramyids were not included in their analyses. Finally, 
any anatomical features of paramyids retained in sciurids are 
symplesiomorphies and not useful for phylogenetic assessments, 
in spite of their interest for other evolutionary studies. 
A number of postcranial features do appear to correlate well 
with terrestrial, arboreal, or gliding habits of Recent squirrels. 
We consider the function of these features, in the hope that there 
are reasons to consider them functional correlates, independent 
of lineage. If so, then these features can better be argued to be 
indicative of these habits, even among fossil squirrels. 
The robust deltoid ridge of the humerus in ground squirrels 
and tree squirrels serves for the insertion of the pectoralis and 
deltoid muscles. The strong distal insertion of these muscles on 
the humerus increases the power of forelimb flexion but reduces 
flexibility at the shoulder joint. The less robust deltoid ridge in 
flying squirrels probably allows increased flexibility at the shoul- 
der joint for gliding. A similar but less extreme difference can be 
seen in a comparison of the non-gliding marsupial Gymnobeli- 
deus and the closely related gliding marsupial, Petaurus, of Aus- 
tralia. Therefore we submit that this difference will distinguish 
flying squirrels in the fossil record as well, although we cannot 
assess how early in their history it will be evident. 
In flying squirrels, the reduction of the lateral epicondylar 
ridge must represent a reduction in the leverage of the extensor 
muscles for flexion of the elbow joint, with a concomitant in- 
crease in the ease with which the elbow joint can be extended. 
This might be expected of a gliding mammal, so we find the 
strong lateral epicondylar ridge of Eoglaucomys to be a surprise, 
in contrast with the reduced ridge on all other flying squirrels we 
have studied. 
The third trochanter of the femur may be analogous. The 
reduction of the trochanter and the trochanteric ridge in flying 
squirrels reflects a reduction in the muscle mass and power of the 
superficial gluteal muscle. However, it is not obvious that this 
would have much effect on flexibility at the hip joint and thus a 
more plausible explanation for this universal characteristic of 
Recent flying squirrels should be sought. There is a similar re- 
duction in the size of the insertion of this muscle in the gliding 
marsupial Petaurus. 
Pronation and supination at the wrist joint is reduced or absent 
in flying squirrels. The pronator quadratus muscle is absent in 
many species, and the pisiform bone articulates with both the 
scapholunate and the triquetral bones, probably to increase sta- 
bility at the wrist joint for control during gliding (Thorington, 
1984; Thorington et al., 1998; Thorington and Darrow, 2000). 
Distinguishing morphological features that are associated with 
these changes are the reduction of the distal end of the ulna, the 
absence of the ridge on the ulna for the origin of the pronator 
quadratus, the shape of the pisiform bone, and the articular facet 
on the scapholunate for the pisiform. Among fossil sciurids, 
these features should distinguish flying squirrels from all others. 
In the hind limb, there are a number of features that distin- 
guish ground squirrels from tree squirrels and flying squirrels. 
Some of these can plausibly be linked to the digging activities 
and burrowing habits of ground squirrels, so the presence of 
these characters can be used to argue that a fossil was not a flying 
or tree squirrel. These include the orientation of the lesser tro- 
chanter, which is directed posteriorly and medially in ground 
squirrels. This trochanter is the point of insertion of the psoas 
major and iliacus muscles, which effect the recovery phase or 
protraction of the femur. In tree and flying squirrels the tip of the 
lesser trochanter lies medial to the axis of rotation of the femur, 
taken as a line between the head and the medial condyle of the 
femur, but lateral to this axis in ground squirrels. Accordingly, 
these muscles should provide for greater lateral rotation of the 
femur in tree and flying squirrels and greater medial rotation in 
960 JOURNAL OF VERTEBRATE PALEONTOLOGY, VOL. 25, NO. 4, 2005 
ground squirrels. It is plausible that this contributes to the toes- 
out posture of tree-climbing squirrels but a more toes-in posture 
for burrowing ground squirrels. 
The existence of a small tubercle on the anterior surface of the 
distal end of the tibia would be an unambiguous demonstration 
that a fossil was a flying squirrel. This highly diagnostic feature is 
the origin of the tibio-carpalis muscle, which extends along the 
lateral edge of the flight membrane and is found only in flying 
squirrels, although the tubercle is not found in all flying squirrels. 
In flying squirrels the calcaneal surface of the astragalus ex- 
hibits derived features. The two calcaneal articular surfaces are 
enlarged and occlude the sulcus, contrary to the condition seen in 
many other squirrels. The ancestral condition, a prominent sul- 
cus, is known in fossil squirrels (Douglassciurus) and in Asian 
tree squirrels (Emry and Thorington, 1982). The derived mor- 
phology permits greater mobility of the ankle joint and is asso- 
ciated with extreme plantar inversion of the foot in gliding. The 
foot is held so that the sole is parallel to the flight membrane and 
to the slip stream of the air passing over it. This must decrease 
turbulence and drag during gliding and increase the lift/drag ra- 
tio and the ratio of glide distance to vertical drop. Presumably, 
this is a requisite condition for efficient gliding, because the in- 
verted foot is seen during glides in four of the six Recent mam- 
mal groups that glide (flying squirrels, colugos, sugar gliders, and 
greater gliders; adequate photographs of the feather tail glider 
and the scaly tailed flying squirrels are not available). An oc- 
cluded sulcus on the plantar surface of the astragalus is also seen 
in North American tree squirrels, permitting the reversal of the 
foot during headfirst decent of trees. If the North American tree 
squirrels are ancestral to or a sister group of the flying squirrels, 
as is probable, then the occluded sulcus is a synapomorphy of 
these two groups. Although not diagnostic of flying squirrels 
alone, it is probably characteristic of all Recent and fossil flying 
squirrels. If this assessment is correct, then Pratt and Morgan 
(1989) were right to doubt that Petauristodon of the early Mio- 
cene, Florida, was a flying squirrel, because it has a prominent 
astragalar sulcus. The irony would be that specimens of Pe- 
tauristodon were the basis for James' (1963) assessment of the 10 
dental features that distinguish flying squirrels from other squir- 
rels. 
In conclusion, we submit that identifications of flying squirrels 
in the fossil record, based on teeth alone, should be carefully 
reviewed and verified, with particular concern for other squirrels 
with similar dental characteristics. The 'gold standard' for such 
identifications should be associated postcranial remains exhibit- 
ing diagnostic features found only in Recent flying squirrels. 
ACKNOWLEDGMENTS 
We are grateful for encouragement, comments, and pre- 
publication manuscript reviews by Robert Emry, Robert Hoff- 
mann, William Korth, Ken Rose, and the editor Alisa Winkler. 
We also appreciate the editorial assistance provided by Kather- 
ine Ferrell. 
LITERATURE CITED 
Black, C. C. 1972. Holarctic evolution and dispersal of squirrels (Roden- 
tia: Sciuridae). Evolutionary Biology 6:305-322. 
Black, C. C, and K. Kowalski. 1974. The Pliocene and Pleistocene Sciu- 
ridae (Mammalia, Rodentia) from Poland. Acta Zoologica Craco- 
viensia 19:461-497. 
Bouwens, P., and H. de Bruijn. 1986. The flying squirrels Hylopetes and 
Petinomys and their fossil record. Proceedings of the Koninklijke 
Nederlandse Akademie van Wetenschappen, Series B: Physical Sci- 
ences 89:113-123. 
Bryant, M. D. 1945. Phylogeny of Nearctic Sciuridae. The American 
Midland Naturalist 33:257-390. 
Bruijn, H. de, A. J. van der Meulen, and G. Katsikatsos. 1980. The 
mammals from the lower Miocene of Aliveri (Island of Evia, 
Greece). Proceedings of the Koninklijke Nederlandse Akademie 
van Wetenschappen Series B Physical Sciences 83:241-261. 
Bruijn, H. de, and E. Unay. 1989. Petauristinae (Mammalia, Rodentia) 
from the Oligocene of Spain, Belgium, and Turkish Thrace. Series of 
the Natural History Museum, Los Angeles County 33:139-145. 
Chaimanee, Y. 1998. Plio-Pleistocene Rodents of Thailand. Thai Studies 
in Biodiversity 3:1-303. 
Corbet, G. B., and J. E. Hill. 1992. The Mammals of the Indomalayan 
Region: a Systematic Review. Oxford University Press, Oxford, En- 
gland, 488 pp. 
Cuenca Bescos, G, and J. I. Canudo. 1992. Los Sciuridae (Rodentia, 
Mammalia) del Oligoceno inferior de Montafban y Olalla (Teruel, 
Espana). Discusion sobre el origen de los esciiiridos. Boletin de la 
Real Sociedad Espanola de Historia Natural Seccion Geologica 87: 
155-169. 
Daxner-Hock, G., and P. Mein. 1975. Taxonomische probleme um das 
genus Miopetuarista Kretzoi, 1962 (Fam. Sciuridae). Palaontolo- 
gische Zeitschrift 49:75-77. 
Dehm, R. 1962. Altpleistocane Sauger von Schernfeld bei Eichstatt in 
Bayern. Mitteilungen der Bayerischen Staatssammlung fur Palaon- 
tologie und Historische Geologic 2:17-61. 
Emry, R., and W. W. Korth. 1996. A new genus of squirrel (Sciuridae: 
Rodentia) from the Chadronian of western North America. Journal 
of Vertebrate Paleontology 16:775-780. 
Emry, R., and R. W. Thorington, Jr. 1982. Descriptive and comparative 
osteology of the oldest fossil squirrel, Protosciurus (Rodentia: Sciu- 
ridae). Smithsonian Contributions to Paleobiology 47:1-35. 
Engesser, B. 1979. Relationships of some insectivores and rodents from 
the Miocene of North America and Europe. Bulletin of Carnegie 
Museum of Natural History 14:1-58. 
Guilday, J. E. 1962. The Pleistocene local fauna of the Natural Chimneys, 
Augusta County, Virginia. Annals of the Carnegie Museum 36: 
87-122. 
Guilday, J. E., P. S. Martin, and A. D. McCrady. 1964. New Paris No. 4: 
a Late Pleistocene cave deposit in Bedford County, Pennsylvania. 
Bulletin of the National Speleological Society 26:121-189. 
Guilday, J. E., P. W. Parmalee, and H. W. Hamilton. 1977. The Clark's 
Cave bone deposit and the Late Pleistocene paleoecology of the 
Central Appalachian Mountains of West Virginia. Bulletin of the 
Carnegie Museum of Natural History 2:1-88. 
Heissig, K. 1979. Die friihesten Flughornchen und primitive Ailuravinae 
(Rodentia, Mamm.) aus dem siiddeutschen Oligozan. Mitteilungen 
der Bayerischen Staatssammlung fur Palaontologie und Historische 
Geologic 19:139-169. 
Heissig, K. 2003. Origin and early dispersal of the squirrels and their 
relatives. Deinsea 10:277-285. 
Hight, M. E., M. Goodman, and W. Prychodko. 1974. Immunological 
studies of the Sciuridae. Systematic Zoology 23:12-25. 
Hill, J. E. 1960. The Robinson collection of Malaysian mammals. Bulletin 
Raffles Museum 29:1-112. 
Hill, J. E. 1962. Notes on flying squirrels of the genera Pteromyscus, 
Hylopetes and Petinomys. Annals and Magazine of Natural History 
4: 721-738. 
Hoffmann, R. S., C. G Anderson, R. W. Thorington, Jr., and L. R. 
Heaney. 1993. Family Sciuridae; pp. 419-465 in D. E. Wilson and 
D. M. Reeder (eds.), Mammal Species of the World: a Taxonomic 
and Geographic Reference. 2nd edition. Smithsonian Institution 
Press, Washington, D. C. 
James, G. T. 1963. Paleontology and nonmarine stratigraphy of the 
Cuyama Valley Badlands, California. Part I. Geology, faunal inter- 
pretations, and systematic descriptions of Chiroptera, Insectivora, 
and Rodentia. University of California Publications in Geological 
Sciences 45:1-154. 
Major, C. J. F. 1893. On some Miocene squirrels, with remarks on the 
dentition and classification of the Sciurinae. Proceedings of the Gen- 
eral Meetings for Scientific Business of the Zoological Society of 
London 1893:179-215. 
McKenna, M. C. 1962. Eupetaurus and the living petauristine sciurids. 
American Museum Novitates 2104:1-38. 
McKenna, M. C, and M. Bell. 1997. Classification of Mammals Above 
the Species Level. Columbia University Press, New York, 631 pp. 
Mein, P. 1970. Les sciuropteres (Mammalia, Rodentia) neogenes 
d'Europe occidentale. Geobios 3:7-77. 
Mercer, J. M., and V. L. Roth. 2003. The effects of Cenozoic global 
change on squirrel phylogeny. Science 299:1568-1572. 
THORINGTON ET AL.?FLYING SQUIRRELS IN THE FOSSIL RECORD 961 
Mossman, H. W., J. W. Lawlah, and J. A. Bradley. 1932. The male 
reproductive tract of the Sciuridae. American Journal of Anatomy 
51:89-155. 
Pratt, A. E., and G. S. Morgan. 1989. New Sciuridae (Mammalia: Ro- 
dentia) from the early Miocene Thomas Farm local fauna, Florida. 
Journal of Vertebrate Paleontology 9:89-100. 
Qiu, Z. 1981. A new Sciuroptere from Middle Miocene of Linqu, Shan- 
dong. Vertebrata Palasiatica 19:228-238. 
Qiu, Z. 1991. The Neogene mammalian faunas of Ertemte and Harr Obo 
in Inner Mongolia (Nei Mongol), China - 8. Sciuridae (Rodentia). 
Senckenbergiana Lethaea 71:223-255. 
Qiu, Z., and Y. Liu. 1986. The Aragonian vertebrate fauna of Xia- 
caowan, Jiangsu - 5. Sciuridae (Rodentia, Mammalia). Vertebrata 
Palasiatica 24:195-209. 
Simpson, G. G. 1945. The principles of classification and a classification 
of mammals. Bulletin of the American Museum of Natural History 
85:1-350. 
Skwara, T. 1986. A new "flying squirrel" (Rodentia: Sciuridae) from the 
early Miocene of Southwestern Saskatchewan. Journal of Verte- 
brate Paleontology 6:290-294. 
Steppan, S. J., B. L. Storz, and R. S. Hoffmann. 2004. Nuclear DNA 
phytogeny of the squirrels (Mammalia: Rodentia) and the evolution 
of arboreal!ty using c-myc and RAG1. Molecular Phylogenetics and 
Evolution 30:703-719. 
Sulimski, A. 1964. Pliocene Lagomorpha and Rodentia from Weze 1 
(Poland). Acta Palaeontologica Polonica 9:149-222. 
Thorington, R. W., Jr. 1984. Flying squirrels are monophyletic. Science 
225:1048-1050. 
Thorington, R. W., Jr., and K. Darrow. 2000. Anatomy of the squirrel 
wrist: bones, ligaments, and muscles. Journal of Morphology 246: 
85-102. 
Thorington, R. W., Jr., K. Darrow, and C. G. Anderson. 1998. Wingtip 
anatomy and aerodynamics in flying squirrels. Journal of Mammal- 
ogy 79:245-250. 
Thorington, R. W., Jr., D. Pitassy, and S. A. Jansa. 2002. Phylogenies of 
flying squirrels (Pteromyinae). Journal of Mammalian Evolution 9: 
99-135. 
Thorington, R. W., Jr., A. L. Musante, C. G. Anderson, and K. Darrow. 
1996. The validity of three genera of flying squirrels: Eoglaucomys, 
Glaucomys, and Hylopetes. Journal of Mammalogy 77: 69-83. 
Submitted 21 July 2004; accepted 29 March 2005. 
APPENDIX 1 
Listed below are the USNM specimens that were examined for this 
paper. Abbreviations: p, postcrania; s, skull. 
Family Sciuridae 
Subfamily Ratufinae: Indo-Malayan giant squirrels 
Ratufa affinis: 300928 (s) 
KafH/a 6ico/or 277614 (s), 573966 (p) 
Ratufa indica: 322077 (p), 355785 (s) 
Ratufa macroura: 256734 (s) 
Subfamily Sciurinae: Tree squirrels and flying squirrels 
Tribe Sciurini: Holarctic and Neotropical tree squirrels 
Douglassciurus jeffersoni: 243981 (p) 
Microsciurus mimulus: 309030 (s), 309033 (s) 
Microsciurus alfarl: 338138 (p), 338139 (p) 
Scium; caro/inenMs: 348323 (s), 398573 (s), 505573 (p), 527930 (s) 
Sciurus deppei: 244939 (p), 244942 (p) 
Sciurus niger. 38860/4044 (s), 583337 (p) 
Sciurus vulgaris: 152686 (s) 
Tamiasciurus douglasii: 549262 (s), 549263 (s) 
Tamiasciurus hudsonicus: 39012/14913 (s), 564078 (p) 
Tribe Pteromyini: Flying squirrels 
Aeromys tephromelas: 481192 (s), 196743 (p) 
Eoglaucomys fimbriatus: 173362 (s), 173363 (p), 326363 (p), 
353243 (p) 
Eupetaurus cinereus: uncataloged specimen (p) 
Gkucomy^ W,rinw;: 129708 (s), 141952/A49822 (p), 530557 (p), 
551843 (p) 
Glaucomys volans: 293439 (s), 329701 (s), 397077 (p), 505618 (p), 
506226 (s) 
Hylopetes nigripes: 477995 (s) 
Hylopetes phayrei: 260623 (s) 
Iomys horsfieldii: 292654 (p) 
Petaurista petaurista: 197320 (p), 326359 (s) 
Petinomys genibarbis: 488672 (s) 
Petinomys lugens: 252319 (s) 
Petinomys setosus: 488676 (s) 
Petinomys vordermanni: 481148 (s) 
Pteromys volans: 237587 (s) 
Trogopterus xanthipes: 258250 (p) 
Subfamily Callosciurinae: Oriental squirrels 
Callosciurus erythraeus: 255367 (p) 
Callosciurus notatus: 155680 (p), 574899 (p) 
Callosciurus prevostii: 300947 (s), 300949 (s), 574902 (p) 
Dremomys everetti: 292615 (s) 
Funambulus pennanti: 328003 (s) 
Lariscus insignis: 498572 (s) 
Nannosciurus borneanus: 198077 (p) 
Rhinosciurus laticaudatus: 488519 (s) 
Sundasciurus hippurus: 488405 (s) 
Sundasciurus lowii: 396658 (p) 
Subfamily Xerinae: Ground squirrels and African tree squirrels 
Tribe Xerini: African ground squirrels 
Xerus inauris: 368419 (p) 
Xerus rutilus: 484010 (p) 
Tribe Protoxerini: African tree squirrels 
Epixerus ebii wilsoni: 543105 (p) 
Funisciurus anerythrus: 402914 (s), 539401 (p) 
Funisciurus pyrropus: 539425 (p) 
Heliosciurus rufobrachium: 439137 (s), 439165 (s), 539426 (p), 
543110 (p) 
Paraxerus cepapi: 295201 (p) 
Paraxerus palliatus: 182804 (s) 
Paraxerus poensis: 539393 (p), 539394 (p) 
Paraxerus vexillarius: 540771 (s) 
Profo^ent: ^a/zgen: 481824 (p), 539438 (s), 539439 (p), 539440 (s), 
539443 (s) 
Tribe Marmotini: Holarctic ground squirrels 
Ammospermophilus leucurus: 578030 (p) 
Cynomys ludovicianus: 35012 (p), 65030 (s) 
Marmota flaviventris: 94251 (s) 
Marmota monax: 157613 (s), 191374 (p) 
JpennopMMi 6eecAeyi: 44259 (s), 484951 (p), 484954 (p), 484955 (p) 
Spermophilus columbianus: 81930 (s), 275817 (s), 398302 (p). 
398303 (p) 
Spermophilus franklinii: 497840 (s) 
Spermophilus lateralis: 176460 (s) 
Spermophilus mexicanus: 79569 (s) 
Spermophilus tridecemlineatus: 48457 (s), 398258 (p) 
Spermophilus variegatus: 248462 (p), 349320 (p) 
Tamias dorsalis: 501006 (p), 501007 (p) 
Tamias minimus: 397141 (p), 527615 (p) 
Tamias striatus: 349628 (p), 364947 (p)